The goal of this fellowship proposal is to use protein engineering to develop inhibitors of a protease thought to play a central role in the invasion and metastasis of cancerous cells. The macromolecular protease inhibitor ecotin will be engineered for specificity and activity against the cancer-associated serine protease urokinase-type plasminogen activator (uPA). Computer homology and molecular modeling techniques will used to identify regions of ecotin that may be mutated for greater specificity and activity against uPA. These identified residues will be randomized by mutagenesis and DNA shuffling techniques, expressed by phage display, and screened for mutants with increased inhibitory activity. Although a gene therapy approach eventually could be used for delivery of a macromolecular inhibitor to a tumor cell, understanding the mechanism by which macromolecular inhibitors can strongly inhibit proteases will aid in the development of small peptide and non-peptide inhibitors. These small molecule inhibitors can be designed to emulate the interactions of the macromolecular inhibitor and will be developed for use in protease inhibitor cocktails that block primary tumor growth. In addition, ecotin- protease interactions may be utilized to isolate new target proteases for similar approaches in the treatment of cancer.